new surface bounding box optimalizations.

[MFraters]

Now that the radius of the sphere is defined, the surface_bounding_box can be fully precomputed.

[codecov]

Codecov Report

Merging #408 (9e27556) into main (fe7f82f) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #408      +/-   ##
==========================================
- Coverage   99.43%   99.43%   -0.01%     
==========================================
  Files          85       85              
  Lines        5173     5172       -1     
==========================================
- Hits         5144     5143       -1     
  Misses         29       29              

Impacted Files	Coverage Δ
source/features/fault.cc	98.87% <100.00%> (-0.01%)	⬇️
source/features/subducting_plate.cc	99.16% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fe7f82f...9e27556. Read the comment docs.

[github-actions]

Benchmark	Main	Feature	Difference (99.9% CI)
Slab interpolation simple none	1.289 ± 0.003 (s=354)	1.275 ± 0.003 (s=351)	-1.2% .. -1.1%
Slab interpolation curved simple none	1.488 ± 0.030 (s=308)	1.467 ± 0.031 (s=304)	-2.0% .. -0.9%
Spherical slab interpolation simple none	1.668 ± 0.005 (s=276)	1.654 ± 0.007 (s=268)	-0.9% .. -0.7%
Slab interpolation simple curved CMS	1.402 ± 0.026 (s=326)	1.386 ± 0.027 (s=322)	-1.7% .. -0.7%
Spherical slab interpolation simple CMS	2.104 ± 0.023 (s=228)	2.055 ± 0.028 (s=207)	-2.7% .. -1.9%
Spherical fault interpolation simple none	1.728 ± 0.006 (s=262)	1.687 ± 0.005 (s=268)	-2.5% .. -2.3%
Cartesian min max surface	3.921 ± 0.160 (s=125)	3.934 ± 0.154 (s=107)	-1.4% .. +2.1%
Spherical min max surface	10.860 ± 0.313 (s=41)	10.894 ± 0.224 (s=44)	-1.6% .. +2.2%

[MFraters]

@alarshi @bangerth This may be of interest to the both of you. In the benchmarks here it makes a small but (just) significant difference. It also show the newly opened up the possibility of more optimizations as for example proposed in #308, now that the spherical radius is known.